mlsanityData to 7 Oct 2026

Harmonizing output imbalance for defect segmentation on extremely-imbalanced photovoltaic module cells images

arXiv 2211.05295 ↗First posted 10 Nov 2022cs.CV cs.LG eess.IVcited by 0

Not on the map. Why?

Abstract (arXiv)

The continuous development of the photovoltaic (PV) industry has raised high requirements for the quality of monocrystalline of PV module cells. When learning to segment defect regions in PV module cell images, Tiny Hidden Cracks (THC) lead to extremely-imbalanced samples. The ratio of defect pixels to normal pixels can be as low as 1:2000. This extreme imbalance makes it difficult to segment the THC of PV module cells, which is also a challenge for semantic segmentation. To address the problem of segmenting defects on extremely-imbalanced THC data, the paper makes contributions from three aspects: (1) it proposes an explicit measure for output imbalance; (2) it generalizes a distribution-based loss that can handle different types of output imbalances; and (3) it introduces a compound loss with our adaptive hyperparameter selection algorithm that can keep the consistency of training and inference for harmonizing the output imbalance on extremelyimbalanced input data. The proposed method is evaluated on four widely-used deep learning architectures and four datasets with varying degrees of input imbalance. The experimental results show that the proposed method outperforms existing methods.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 55 references that are in the map. First posted on arXiv 10 Nov 2022.

Builds on

23 of 55 references.

References · 23

  1. 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,845
  2. 2016Pyramid Scene Parsing Networkcited by 2,472
  3. 2017Rethinking Atrous Convolution for Semantic Image Segmentationcited by 2,154
  4. 2015SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentationcited by 1,483
  5. 2021UNETR: Transformers for 3D Medical Image Segmentationcited by 954
  6. 2017Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentationscited by 614
  7. 2021DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detectioncited by 510
  8. 2019Deep Multi-modal Object Detection and Semantic Segmentation for Autonomous Driving: Datasets, Methods, and Challengescited by 313
  9. 2017Tversky loss function for image segmentation using 3D fully convolutional deep networkscited by 228
  10. 2020A survey of loss functions for semantic segmentationcited by 216
  11. 2018A Novel Focal Tversky loss function with improved Attention U-Net for lesion segmentationcited by 171
  12. 2019Segmentation-Based Deep-Learning Approach for Surface-Defect Detectioncited by 145
11 more
  1. 2021Inpainting Transformer for Anomaly Detectioncited by 91
  2. 2021Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentationcited by 88
  3. 2018Combo Loss: Handling Input and Output Imbalance in Multi-Organ Segmentationcited by 76
  4. 2022A Survey on Long-Tailed Visual Recognitioncited by 75
  5. 2016Automatic Liver and Lesion Segmentation in CT Using Cascaded Fully Convolutional Neural Networks and 3D Conditional Random Fieldscited by 43
  6. 20183D Segmentation with Exponential Logarithmic Loss for Highly Unbalanced Object Sizescited by 42
  7. 2019A Survey on Deep Learning-based Architectures for Semantic Segmentation on 2D imagescited by 29
  8. 2018Automatic segmentation of the spinal cord and intramedullary multiple sclerosis lesions with convolutional neural networkscited by 14
  9. 2015Beyond Pixels: A Comprehensive Survey from Bottom-up to Semantic Image Segmentation and Cosegmentationcited by 10
  10. 2018Solar Cell Surface Defect Inspection Based on Multispectral Convolutional Neural Networkcited by 6
  11. 2019Fully Convolutional Networks for Chip-wise Defect Detection Employing Photoluminescence Imagescited by 1

Built on by

Not cited yet.